Table 3.

Potential explanations for the apparent discrepancy in clinical response reported in clinical experience and DBRCTs

	Clinical experience	Randomised controlled trials
Disease activity	Refractory to conventional immunosuppressants	Rituximab was used as an add-on therapy to background immunosuppressants
	Favourable response reported in life-threatening cases, often including a range of organ-system involvement such as CNS manifestations, cytopenias and others	Life-threatening cases and those with CNS manifestations were not evaluated in controlled trials. This setting warrants a dedicated study
Clinical response	No defined pretreatment, therefore complete and partial responders might not be clearly distinguished	Predefined endpoints were stringent, perhaps driven by the impressive responses seen in clinical experience in an uncontrolled setting
	Improvement in one system alone might qualify for response, regardless of a flare or lack of response in another organ system	Predefined and usually stringent. For example, despite clinical response and steroid-sparing effect, a reduction in proteinuria that does not meet the predefined threshold would not qualify as complete/partial response
Background immunosuppressants	Flexibility in changes to background immunosuppressants including the dose of corticosteroids	Changes to or deviation with predefined background therapy would qualify as nonresponder
	Concomitant use of large dose of steroids is uncommon	Concomitant use of large dose of corticosteroids might have limited any beneficial effects of rituximab, the extent of which may be more restricted in such a setting than previously assumed
Rituximab dosing-regimen	Variable between reports	Predefined dosing regimen
Steroid tapering	Steroid-sparing effect is not a requirement to define response and therefore favourable response might be overestimated	Steroid dosing effect was included in the definition of clinical response
Adverse events	No standardised reporting of adverse events. Therefore, the true incidence of serious adverse events in clinical practice is not comparable with that reported in other uncontrolled studies or controlled clinical trials	Rituximab therapy appears to be safe as no there were no significant differences in serious adverse events when compared with standard-of-care treatment
Follow-up period	Not defined, therefore it is not known how many responders had sustained response in the long term	Predefined, therefore, unless long-term studies are undertaken, it would be difficult to detect the importance of effects seen at relatively short-term follow-up

CNS, central nervous system.